在线社交网络中基于多态信任融合的信任估计
Trust estimation based on polymorphic trust fusion in online social networks
任蓓蓓 1陈玉泉 2陈芸2
作者信息
- 1. 上海公安学院信息化与网络安全系,上海 200137
- 2. 上海交通大学计算机科学与工程系,上海 200052
- 折叠
摘要
为提高大型在线社交网络中信任计算方法的准确性和鲁棒性,采用共被引和转置信任传播操作,提出一种基于多态信任融合的信任估计方法估计连续信任/不信任值.结合信任者、被信任者的相邻用户的信息以及被信任者对信任者的信任,平均估计出两个用户信任或被其它用户信任的差异,以及一个用户信任另一个用户和被该用户信任的差异;利用这些差异,计算4种部分信任估计值,将这些部分估计值加权平均,得到信任者对被信任者的最终信任估计值.仿真结果表明,所提方法比其它最新的现有信任计算算法更准确和鲁棒,对应用于大型网络更高效.
Abstract
To improve the accuracy and robustness of trust calculation methods in large online social networks,a trust estimation method based on polymorphic trust fusion to estimate continuous trust/distrust values was proposed using co-citation and trans-pose trust propagation operations.To solve this problem,by combining the information from neighboring users of trustor,trus-tee and the trust from the trustee to trustor,the difference between two users'trust or trust trusted by other users,and the difference between one user trusting in another user and trust trusted by that user were estimated on average.Using these diffe-rences,four partial trust estimates were calculated,and these partial trust estimates were weighted average to get the final trust estimate from trustor to trustee.Simulation experiment results show that the proposed method is more accurate and robust than other latest existing trust algorithms,and is more efficient for large networks.
关键词
在线社交网络/信任传播操作/信任计算/加权有向图/部分信任估计/均方根误差/鲁棒性Key words
online social network/trust propagation operation/trust computing/weighted directed graph/partial trust estima-tion/root mean square error/robustness引用本文复制引用
出版年
2024
NETL
NSTL
万方数据